Fast and highly accurate (99.9% accuracy for most major languages*) language detection in 176 languages based on the Fast Text Machine Learning model. Please see the “About” section for details on languages and accuracy data.

99% Accurate Languages*: French (fr), English (en), German (de), Portuguese (pt), Turkish (tr), Dutch (nl), Italian (it), Spanish (es), Hungarian (hu), Esperanto (eo), Polish (pl), Finnish (fi), Russian (ru), Macedonian (mk), Ukrainian (uk), Lithuanian (lt), Vietnamese (vi), Greek (el), Marathi (mr), Arabic (ar), Hebrew (he), Hindi (hi), Uyghur (ug), Japanese (ja), Georgian (ka), Bengali (bn), Urdu (ur), Thai (th), Chinese (zh), Armenian (hy), Malayalam (ml), Korean (ko), Khmer (km), Burmese (my), Tamil (ta), Kannada (kn), Telugu (te), Panjabi (pa), Lao (lo), Gujarati (gu), Tibetan Standard (bo), Divehi (dv), Sinhala (si), Amharic (am).

90% Accurate Languages*: Danish (da), Romanian (ro), Swedish (sv), Latin (la), Bulgarian (bg), Czech (cs), Tagalog (tl), Indonesian (id), Tatar (tt), Icelandic (is), Belarusian (be), Basque (eu), Breton (br), Kazakh (kk), Latvian (lv), Estonian (et), Irish (ga), Chuvash (cv), Bashkir (ba), Ossetian (os), Tajik (tg).

*Information is presented in order of most test data. Data was sentences of 30-250 characters in length. Testing was only done on the most popular 100 languages. Testing showed near a 99% accuracy for majority of sentences at or above 300 characters in length.